Efficient determination of optimal radial power system structure using Hopfield neural network with constrained noise

When a radial power system has a number of connected feeders, the total number of possible system structures can be very large. In order to determine the optimal radial power system structure rapidly, we propose a constrained noise approach, which can avoid local minima, with the Hopfield neural network model. For checking the validity of the proposed approach we compare the proposed method with a conventional branch-and-bound method which is popular in the field of mathematical programming. Simulations are carried out for two actual subsystems of Tokyo Electric Power Co. (TEPCO). Furthermore, because engineering knowledge is necessary to operate or plan the radial power system securely, we combine the proposed Hopfield model with engineering knowledge in order to obtain a more practical system structure considering cases of fault occurrence at each substation. The combined technique is demonstrated with one of the TEPCO subsystems.